Does the dividend/price ratio forecast future dividend movements as re-
quired by the random-walk theory, or does it instead forecast future move-
ments in stock prices? We answer this question using a long-run annual
U.S. data set that extends today’s S&P 500 Index back in time to 1872.^3
The answer is given by the pair of scatterplots shown in figure 5.1. Each
scatterplot has the dividend/price ratio, measured as the previous year’s div-
idend divided by the January stock price, on the horizontal axis. (The hori-
zontal axis scale is logarithmic but the axis is labeled in levels for ease of
reference.) Over this period the historical mean value for the dividend/price
ratio was 4.65 percent.
In the top part of the figure the vertical axis is the growth rate of real div-
idends (measured logarithmically as the change in the natural log of real
dividends) over a time interval sufficient to bring the dividend/price ratio
back to its historical mean of 4.65 percent. More precisely, we measure the
dividend growth rate from the year preceding the year shown until the year
before the dividend/price ratio again crossed 4.65 percent. Because divi-
dends enter the dividend/price ratio with a one-year lag, this is the appro-
priate way to measure growth in dividends from the base level embodied in
a given year’s dividend/price ratio to the level that prevailed when the divi-
dend/price ratio next crossed its historical mean.^4
Since 1872, the dividend/price ratio has crossed its mean value twenty-
nine times, with intervals between crossings ranging from one year to
twenty years (the twenty-year interval being between 1955 and 1975). Se-
lected years are indicated on the scatter diagram by two-digit numbers; a *
after a number denotes a nineteenth-century date. The last year shown is
1983, since this is the last year that was followed by the dividend/price
ratio crossing its mean. (The ratio has been below its mean ever since.) A
regression line is fit through these data points, and a vertical line is drawn
to indicate the dividend/price ratio at the start of the year 2000. The im-
plied forecast for dividend growth is the horizontal dashed line marked
where the vertical line intersects the regression line.
VALUATION RATIOS 175(1988a) paper for a careful analysis of dividend forecasts within the context of a log-linearized
mathematical representation of the efficient-markets theory, or Campbell, Lo, and MacKinlay
(1997), chapter 7, for a recent textbook exposition.
(^3) The data in this chapter use the January S&P Composite stock price for each year since
1872, while earnings and dividends are for the entire previous year. Data before 1926 are based
on Cowles (1939). The price index used to deflate nominal values to real values is the pro-
ducer price index. See Shiller (1989) for a description of these data.
(^4) The time intervals required to bring the dividend/price ratio back to its mean typically ex-
ceed one year, so the dividend growth rate for any particular year can affect several successive
observations. This overlapping of successive time intervals implies that the different points in
the scatterplot are not statistically independent. There are, however, twenty-nine nonoverlap-
ping time intervals in our sample, so the data are not insubstantial. Statistical tests of the signif-
icance of analogous relations with fixed horizons, taking account of the overlapping intervals,
are reported in Campbell and Shiller (1988b, 1989).